Interactions with soil microbes can strongly affect plant growth and defense against aboveground herbivores. Plant species often accumulate specific soil pathogens in their rhizosphere, leading to reduced growth of plants in soils originating from stands of conspecific plants compared to soils from heterospecific plants. However, whereas effects of such conspecific vs. heterospecific soil biota on plant growth have been well documented, their effects on plant resistance and tolerance to aboveground insect herbivores have not. We compared growth and defense of Triadica sebifera plants from populations where the species is native (China), when grown in sterilized soils, or in soils harbouring belowground biota from conspecific (native Triadica) or heterospecific (native grass) soils. In each of these soils, plants were exposed to a 15-day period of foliar herbivory by a specialist weevil (Heterapoderopsis bicallosicollis), a specialist caterpillar (Gadirtha inexacta), or no herbivory (cage), followed by a 60-day recovery period. Soil biota from conspecific and hetetospecific soils differed in their effects on plant growth and defense. First, in the absence of herbivory, soil biota from heterospecific soils slightly enhanced plant growth, whereas those from conspecific soils strongly reduced plant growth. Second, soil biota from conspecific soils strongly affected plant resistance and tolerance to foliar herbivory, whereas soil biota from heterospecific plants did not. The effects of soil biota on plant defense were herbivore-specific. In particular, conspecific soil biota reduced resistance to caterpillar but not to weevil feeding, whereas they enhanced tolerance to weevil but not to caterpillar feeding. Conspecific soil biota also mitigated induction of root flavonoids by herbivores and led to reduced root phenolics in response to herbivory. Conversely, caterpillar feeding increased AMF colonization, but under these conditions, AMF colonization was negatively associated with plant biomass. In addition to testing effects on native plants, we also tested effects of native soil biota on growth and resistance of plants from the introduced range (North America). Plants from the introduced range had higher shoot production, shoot-to-root ratio, and leaf phenolic and flavonoid production than plants from the native range, but their interactions with soil biota showed only minor differences compared to plants from the native range. Our results suggest that incorporating the effects of soil biota in interactions between plants and foliar herbivores is critical for understanding variations in growth, defense, and performance among plant populations at local and broader geographic scales.